Why not use mains as centers


Is there a reason I should'nt use two good bookshelfs or smaller towers for my center if I have two center outputs and available amp channel?
rmichael21
Well, gentlemen, you can disagree with me all you want but the physics does not change. Two drivers emitting the same signals but displaced from each other by more than 1/2 wavelength will produce additive and subtractive interactions resulting in an extremely ragged spatial radiation pattern in the plane of the drivers. This means, for the typical center channel, in the horizontal plane, unfortunately.

Thus, a typical MTM arrangement is fine when vertical since its horizontal dispersion is uniform and the vertical dispersion less important as long as the listeners remain seated or in the direct beam of the speaker. A "center channel with 2 mid drivers flanking the right and left side of a tweeter" is a design that is made to accommodate design and convenience and suffers from the described uneven dispersion. Note the lack of any quality main speakers with laterally arrayed drivers.

I do not wish to single out ML, nht or any other decent company since they are almost all at fault. They want to sell speakers that people will want and, unfortunately, most people think that a wide center speaker with an array of drivers "looks" like it would have a wide dispersion.

None of what I say depends on my hearing but on physics. Speak with any design engineer or read any book on speaker engineering and the issue is clear. Whether you care about it is your own business.

Kal
Jaybo wrote: "two centers as mains(l and r) will work too."

If you stand them up vertically. ;-)

Kal
Kal,

You are quite correct. Speaker arrays (same drivers outputting same frequencies) are best avoided except in the bottom two octaves (assuming one cherishes precision). You get lobes in response as well as comb filtering of certain frequencies which makes the audible response much more variable than a point source.
As I stated earlier, the issue is the relationship of the driver spacing to the wavelength. In the bottom end, the wavelengths are large enough that the relative placement of the drivers is inconsequential. I was surprised at the resistance to this issue here as I thought it was a fairly well-known matter.

Kal